* Update PENDING.md * Add simple period for expensive invariants * Remove individual module simulations * Simulate a few more blocks * Add README explaining reason for shell scripts * Deduplicate scripts, log exact replication command on failure * Refactor invariants to take sdk.Context instead of baseapp.BaseApp * Reference all issues in PENDING.md entry * Remove no longer used simulation.RandSetup * Bug fixes * Address @rigelrozanski comments * Fix typo |
||
---|---|---|
.. | ||
install | ||
Makefile | ||
README.md | ||
linkify_changelog.py | ||
localnet-blocks-test.sh | ||
multisim.sh |
README.md
Generally we should avoid shell scripting and write tests purely in Golang. However, some libraries are not Goroutine-safe (e.g. app simulations cannot be run safely in parallel), and OS-native threading may be more efficient for many parallel simulations, so we use shell scripts here.